At present, how to make the computer understand the text message of humanity automatically is a very important issue in computer information technology field. And the problem of word sense disambiguation is a bottleneck of the understanding of natural language. This paper describes the current research situation of word sense disambiguation, introducing its background and application, analyzing various mature approaches, comparing the merits and shortages of those methods, taking a brief introduction of context and international evaluation method of word sense disambiguation and finally making a conclusion of its future and development trend.